• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Akai GX-620 Reel to Reel Tape Deck Measurements

Stephen H

Member
Joined
Oct 23, 2022
Messages
21
Likes
141
I recently had the idea to use my hi-spec ADC in combination with a USB sound card to have a go at creating my own measurements. This is the first time I've done this, so keep in mind there may be some mistakes etc. This is not my area of expertise.

PXL_20221023_012028367.jpg


Above is my beautiful Akai GX-620.

I replaced all the dodgy Akai transistors and swapped out the electrolytic capacitors whilst I was at it. I have always been quite surprised by how good this thing sounds when listening to the source. However, I do find there is a bit too much bass when recording to tape. This could be caused by many things. I don't imagine it is particularly well calibrated, and the tape heads have probably seen better days. However, I've wanted to have a go at testing and measuring something for a while now, and this felt like the perfect starting point.

Devices used for capture:
  • Desktop PC running Windows 11.
  • SMSL SU-8 v2 DAC.
  • Nihtila ADC AK5572
  • SoundBlaster SB1240
  • Multitone software.
Chain of devices (in loop): Windows PC -> SMSL DAC -> Akai GX-620 -> Nihtila ADC -> SoundBlaster 1240 (via Toslink) -> Windows PC

Documentation & Notes:

  • The SMSL DAC has already been reviewed and measured courtesy of our friend Amir: Click Here
  • Audio Precision results for the ADC can be found on Tomi Nihitila's website: Click Here
  • Toslink connects from the ADC to SoundBlaster 1240.
  • SoundBlaster Connects to PC via USB.
  • SoundBlaster responsible for Digital-to-Digital conversion only.
  • Bitrates all set to match between DAC and ADC etc (96khz).
  • Measurements have not been offset to 0db.
  • ADC is being used single-ended and not balanced which loses 6db of dynamic range.
The below measurements shows the SMSL DAC plugged directly into ADC:

1Khz Tone:

SU-8 1khz.png


I was very surprised to see how similar it is to Amir's review. It is more or less identical. So nothing much to see here!

Moving on to a 32-tone signal:

SU-8 Multitone.png


Seems to be behaving as I would expect. Although, one thing to keep in mind, the above is not offset to 0db.

Frequency Response:

SU-8 FQ.png


Nothing exciting to see here. Good and flat as you would expect.

Moving on to something more exciting... the below measurements show the tape deck hooked up between the DAC and the ADC.

First I measured the pre-amp alone (not recording).

1Khz Tone:

Akai GX-620 1Khz.png


The above result was better than I was expecting! I reduced the voltage down slightly to not overload the pre-amp board, hence why the fundamental is 10db down. Also bit of mains noise going on (50hz here in the UK) and some harmonic distortion.

32-tone signal:

Akai GX-620 Multitone.png


Doesn't seem too bad, noise floor would be much higher (almost 30db) if you were to offset the results for calculating SNR etc. Looks like the same mains noise going on.

Frequency response:

Akai GX-620 FQ.png


Seems flat enough to me. Little roll-off either end, but nothing major wrong here.

Now, the really interesting part! Let's record these tones to tape and see what happens. I used SM911 tape from Thomann.de

1Khz Tone:

Akai GX-620 1Khz Tape.png


Harmonic spikes have increased in size. Noise floor has also raised due to tape hiss. Personally, I am still pleasantly surprised by this result. I was expecting to see absolute carnage!

32-tone signal:

Akai GX-620 Multitone Tape.png


I'm afraid you guys are more or less on your own with the above result here. There is seemingly lots of distortion and the noise floor is uneven around 50-100hz. Also, you can see the frequency response is looking a little sloped down from the bass side of things. More below.

Frequency response:

Akai GX-620 FQ Tape.png


Yeah, this is pretty ugly compared to what we're used to seeing on this website. As I have always thought, there is elevation in low frequencies. There is also a bulge around 50hz which I'm guessing it due to mains in the UK. I wonder if this would show any improvement with re-capping the power supply board? I have only re-capped the pre-amp board so far.

All in all, I was surprised by the above. I was expecting it to be a total horror show. Apart from the uneven frequency response, it doesn't seem too bad. However, as I first mentioned, I am not an expert in this area by any stretch of the imagination. I am merely a hobbyist.

I hope anyone who took the time to read this post has managed to get something out of it. Unfortunately, I won't be the best at explaining what it all means. However, I will say that, I was always surprised by how good this deck sounds. Obviously it doesn't sound like a modern day DAC, but I would never expect it too. I always had noticed the recordings from the deck had more bass, and sounded a little muddy when compared to the source. It's nice to see the data back this up by showing the uneven frequency response of the recording when compared to the pre-amp only measurement. Unsurprisingly, apart from the SU-8 alone, none of the above measurements look like what we're used to seeing here.

As a small side note. I can highly recommend the AK5572 ADC. What a fantastic little board that is. Coupled with the SPDIF converter here, the result is a truly fantastic performing ADC once boxed and fed with a decent PSU. I cannot remember exactly what PSU I used now, however it was a small silent-switcher style design. Clearly the above demonstrates that all working too!
 
Last edited:

mhardy6647

Grand Contributor
Joined
Dec 12, 2019
Messages
11,217
Likes
24,181
What tape type did you use and was the deck adjusted for it (bias & EQ)? Looks like no given the "uncalibrated" comment in the OP(?).
What tape speed did you use?
What recording level did you use?

EDIT: I'm sorry, that was way too brusque! :(
Great first effort and an interesting place to start! Nice photo, too!
I suspect (?!?) there may be too many variables to go too deep in terms of interpreting the test recording data, though.
 
Last edited:

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,580
Likes
38,281
Location
Gold Coast, Queensland, Australia
My favourite RTR, the Akai GX-625!

Basically identical except a few more controls (pitch, timer and a cool digital real-time memory counter with associated control circuitry)

IMG_0797 (Medium).jpg
IMG_0800 (Medium).jpg
IMG_0801 (Medium).jpg


All I gotta do is find the NOS reels of EMTEC (BASF) 468 I have someplace...

But you are right, the Akai has a very 'warm' bottom end for sure. Compared to my other open reels, it sounded fat and fabulous, so much so, I sold off most of the others. If you want open reel warmth, these decks offer it up in spades. Sure the Pioneer RT-909 might have been prettier, but it was a pain to keep in top condition, whereas the Akais with the eddy current outer rotor motors are sweet in that regard. The DD capstan motor and the original pinch rollers last too. Glass crystal heads are practically indestructible too.

And get rid of those 'tombstone' Hitachi transistors!
 
Last edited:

respice finem

Major Contributor
Joined
Feb 1, 2021
Messages
1,867
Likes
3,774
Oh those memories...
Could worn (or not demagnetized) heads be the main culprit?
They should last "forever", but theory is theory and life is life...
 

JaccoW

Senior Member
Joined
Sep 25, 2021
Messages
348
Likes
512
Location
The Netherlands

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,580
Likes
38,281
Location
Gold Coast, Queensland, Australia
Oh those memories...
Could worn (or not demagnetized) heads be the main culprit?
They should last "forever", but theory is theory and life is life...

I think it’s more likely the bias/eq not being suitable for the tape used and possibly some head wear/azimuth issues. Easy to do azimuth with a reference tape, but classic circle of confusion without.

If the OP sticks to one formulation of tape, he can calibrate the deck to that and the response can be optimised.
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,580
Likes
38,281
Location
Gold Coast, Queensland, Australia
I've always loved the light-up buttons on these decks. Not that I'm unhappy with my Akai GX-635D! :D




I'm currently redoing my setup and I've got a Focusrite Scarlett 8i6 so perhaps I'll try my hand at some measurements as well. ;)

Do it! If silly me hadn’t packed all my RTR stuff (nab adapters, demags, ref tapes and NOS 10.5”) stuff into my storage, I’d do it too.
 

eeMGee

Member
Joined
Sep 22, 2022
Messages
25
Likes
11
Very nice, but I suggest to switch level in your charts to dBr (not dBFS) - Multitone Analyzer calculates THD relative to FS, therefore when your fundamental is lower than 0 dBFS, THD is not correct.
In your case is THD about 10 dB lower then in reality...
 
Last edited:

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
Perhaps redo the measurements with the other tape type.
Also do the FR measurents at -10dB and/or -20dB this might give much better (realistic) FR measurements of course at the cost of S/N ratio.
 
OP
Stephen H

Stephen H

Member
Joined
Oct 23, 2022
Messages
21
Likes
141
What tape type did you use and was the deck adjusted for it (bias & EQ)? Looks like no given the "uncalibrated" comment in the OP(?).
What tape speed did you use?
What recording level did you use?

EDIT: I'm sorry, that was way too brusque! :(
Great first effort and an interesting place to start! Nice photo, too!
I suspect (?!?) there may be too many variables to go too deep in terms of interpreting the test recording data, though.
No probs :)

Tape speed = 7.5ips.

Recording level set at 0db (deck runs at 250nWb/m reference fluxivity).

Indeed a lot of variables here! Far too many for me to investigate. The deck could certainly be tweaked and further calibrated to improve performance. However, it's good enough for me!
 
OP
Stephen H

Stephen H

Member
Joined
Oct 23, 2022
Messages
21
Likes
141
My favourite RTR, the Akai GX-625!

Basically identical except a few more controls (pitch, timer and a cool digital real-time memory counter with associated control circuitry)

View attachment 239080View attachment 239081View attachment 239082

All I gotta do is find the NOS reels of EMTEC (BASF) 468 I have someplace...

But you are right, the Akai has a very 'warm' bottom end for sure. Compared to my other open reels, it sounded fat and fabulous, so much so, I sold off most of the others. If you want open reel warmth, these decks offer it up in spades. Sure the Pioneer RT-900 might have been prettier, but it was a pain to keep in top condition, whereas the Akais with the eddy current outer rotor motors are sweet in that regard. The DD capstan motor and the original pinch rollers last too. Glass crystal heads are practically indestructible too.

And get rid of those 'tombstone' Hitachi transistors!

Beautiful machine! I love the digital counter on the 625!

Interesting to know that you also find it to has a warm sounding low end.

I can see how this could be appealing to a lot of people.

Definitely have to replace the transistors, they cause all sorts of horrible noise and distortion. You can physically see the legs on the old transistors corrode.
 
OP
Stephen H

Stephen H

Member
Joined
Oct 23, 2022
Messages
21
Likes
141
Perhaps redo the measurements with the other tape type.
Also do the FR measurents at -10dB and/or -20dB this might give much better (realistic) FR measurements of course at the cost of S/N ratio.
I did try both tape types. Made more or less no difference to measurements!
 
OP
Stephen H

Stephen H

Member
Joined
Oct 23, 2022
Messages
21
Likes
141
Very nice, but I suggest to switch level in your charts to dBr (not dBFS) - Multitone Analyzer calculates THD relative to FS, therefore when your fundamental is lower than 0 dBFS, THD is not correct.
In your case is THD about 10 dB lower then in reality...
Thanks for the tip! I will keep that in mind.
 
OP
Stephen H

Stephen H

Member
Joined
Oct 23, 2022
Messages
21
Likes
141
I've always loved the light-up buttons on these decks. Not that I'm unhappy with my Akai GX-635D! :D




I'm currently redoing my setup and I've got a Focusrite Scarlett 8i6 so perhaps I'll try my hand at some measurements as well. ;)
Sorry deleted my previous reply by accident. You should definitely have a go at measuring it if you get time! I am highly jealous of the each-way play feature :D
 
OP
Stephen H

Stephen H

Member
Joined
Oct 23, 2022
Messages
21
Likes
141
Oh those memories...
Could worn (or not demagnetized) heads be the main culprit?
They should last "forever", but theory is theory and life is life...
Maybe sometime in the future I'll mess around with the deck and try and improve head-alignment/calibration etc. For now, I am happy enough with it
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,580
Likes
38,281
Location
Gold Coast, Queensland, Australia
Maybe sometime in the future I'll mess around with the deck and try and improve head-alignment/calibration etc. For now, I am happy enough with it

Leave it until you have access to an actual genuine calibration tape. Not a tape someone 'made' for you or one you 'made' yourself...
 
Top Bottom